CCP4 Interface: Tutorial 2

	CCP4i: Graphical User Interface
	Tutorial 2

Molecular replacement

AMoRe - more involved examples

The correlation coefficients and Rfactors as they came out of the simple AMoRe example are not particularly promising. It is, therefore, wise to play around with some of the parameters.

Do not worry about the jumping of windows. Whenever only a shape of a window comes up, position it on the screen where you want it (as before).

Varying the search radius

The Interface calculates an optimal value for the search radius (see AMoRe Rotation Function Radius and Model Cell): 19.396 in this case. Opinions differ on how crucial this number actually is (see AMoRe Program Documentation). The larger it is, the more chance there is of including some Patterson vectors twice. It is also thought that the volume of the sphere should probably not exceed the volume of the asymmetric unit. The program allows a margin in this, but stops when the difference in volume gets too large: if the sphere volume is larger than 1.2*asymmetric unit volume, the program gives a friendly warning; if the sphere volume is larger than 1.5*asymmetric unit volume, the program stops with a fatal error warning.

The parameter to change in the Interface, is
Rotation function search sphere in the 'Key Parameters' folder.
It would also be wise to increase the number of peaks output by decreasing the threshold fraction of maximum peak height to 0.3 both in the Rotation Function and Translation Function Parameters folders. Find 8 peaks above this fraction is probably enough to get an idea of the influence of noise peaks.

Evaluation

Results of several runs of auto-AMoRe are summarised in the following table. The resolution range is kept as 10-3Å throughout. The model is the same as before.

sphere radius	ROTFUN					TRAFUN				FITFUN			comments
sphere radius	peak nr	CC_F	RF_F	CC_I	CC_P	peak nr	CC_F	RF_F	CC_I	CC_F	RF_F	CC_I	comments
increasing from optimal value
20	1	22.6	53.0	35.1	29.1	2	71.2	33.9	70.0	83.2	26.8	83.2	program warns that the sphere radius is larger than 0.5*cell_edge
	2	23.4	52.8	36.3	29.1	1	74.7	31.9	72.9	83.2	26.8	83.2
	3	8.2	56.9	13.7	11.7	7	13.9	56.5	14.2	24.1	54.9	24.1
	4	9.6	56.6	14.3	11.6	3	21.0	54.3	19.3	30.3	53.0	27.9
22	1	22.6	53.0	35.1	26.9	2	71.3	33.8	70.1	83.2	26.8	83.2	Only 7 ROTFUN solutions higher than half maximum
	2	23.1	52.9	35.8	26.9	1	73.4	32.7	71.8	83.2	26.8	83.2
	3	9.4	56.6	14.2	9.7	3	20.1	54.3	18.7	29.6	53.1	27.3
	4	9.7	56.7	13.1	9.4	6	15.0	56.7	14.6	25.0	55.3	24.1
24	1	23.7	52.7	36.7	25.6	2	76.0	31.1	74.0	83.2	26.8	83.2	Only 5 ROTFUN solutions higher than half maximum
	2	23.8	52.6	36.8	25.6	1	76.3	30.9	74.3	83.2	26.8	83.2
	3	9.5	56.6	14.3	8.5	3	20.6	54.2	18.8	29.8	52.7	27.0
	4	7.3	57.5	12.1	8.0	5	12.9	56.1	13.8	23.7	54.8	24.7
26	1	24.1	52.5	37.0	24.3	2	77.5	30.0	75.3	83.2	26.8	83.2	Only 5 ROTFUN solutions higher than half maximum
	2	24.1	52.5	37.0	24.3	1	77.7	29.9	75.4	83.2	26.8	83.2
	3	9.3	56.9	14.0	7.4	4	16.9	55.5	16.9	27.4	53.9	26.5
	4	9.7	56.8	14.6	7.4	3	20.4	54.1	18.7	29.6	52.8	27.0
26.1	program stops with fatal error because of the large sphere volume
decreasing from optimal value
18	1	22.4	53.1	34.9	29.5	2	70.5	34.3	69.5	83.2	26.8	83.2
	2	23.5	52.8	36.3	29.5	1	74.9	31.8	73.1	83.2	26.8	83.2
	3	9.7	56.7	14.4	13.5	3	20.0	54.3	18.8	29.5	52.8	27.5
	7	9.7	56.7	14.6	11.4	4	15.9	55.7	15.5	26.5	54.1	25.7
16	1	23.1	52.9	35.8	31.0	2	73.5	32.6	72.0	83.2	26.8	83.2
	2	23.7	52.6	36.7	31.0	1	76.1	31.0	74.1	83.2	26.8	83.2
	3	9.0	56.7	13.6	15.0	3	18.6	54.4	17.3	28.8	52.6	26.9
	5	9.7	56.7	14.5	13.7	4	15.9	55.6	16.2	26.1	54.2	26.3
14	1	22.4	53.1	34.8	31.7	2	70.2	34.5	69.2	83.2	26.8	83.2
	2	23.0	52.9	35.7	31.7	1	73.1	32.8	71.7	83.2	26.8	83.2
	3	8.8	56.8	15.3	16.9	4	14.1	56.6	13.1	24.3	54.8	22.7
	4	10.0	56.8	15.3	16.9	3	20.0	54.5	18.1	19.4	53.3	27.1
12	1	24.1	52.5	37.0	32.5	2	77.6	29.9	75.4	83.2	26.8	83.2
	2	24.1	52.5	37.0	32.5	1	78.0	29.6	75.6	83.2	26.8	83.2
	3	7.7	57.4	10.6	23.1	8	13.5	56.7	12.3	24.4	54.7	22.1
	4	10.3	56.6	16.0	19.9	3	18.0	55.5	18.2	30.5	52.7	29.8
10	1	18.6	54.1	29.9	31.3	1	52.2	42.9	54.4	83.2	26.8	83.2
	2	17.5	54.4	28.3	31.3	2	47.5	44.8	50.0	83.2	26.8	83.2
	3	7.7	57.3	11.4	30.0	5	13.3	56.6.	12.4	23.0	55.3	21.8
	4	7.7	57.5	11.1	24.1	3	13.9	56.1	12.5	24.7	54.3	22.5
8	1	9.7	56.8	15.7	42.1	5	15.7	56.0	15.6	25.7	54.5	24.6
	2	9.5	56.7	15.2	41.5	1	16.7	55.5	17.4	81.1	28.4	81.7
	3	as peak 2				2	as above			81.3	28.3	82.8
	4	as above				3	as above			81.1	28.4	81.7
	5	as above				4	as above			81.2	28.4	81.8

N.B. for most of the near-optimal choices of the sphere radius, the highest noise peak is less than half the height of the maximum peak (i.e. the correct peak).

Conclusions

Increasing the sphere radius only becomes a problem when the sphere volume becomes excessively large compared to the volume of the asymmetric unit and the program stops. Apart from that, the correct solutions are found easily.
Decreasing the sphere radius mainly increases the number of peaks above half the maximum peak height. At some point, the correct peaks come out lower than the highest noise peak at the ROTFUN stage. This is fortuitously 'rectified' at the FITFUN stage. It does, however, not seem sensible to use a sphere radius smaller than half the optimal value.
In the case of RNAse SA, the choice of sphere radius does not seem to be crucial. It may, however, be quite different for other molecular replacement problems.

More information

For more information on the AMoRe program, see AMoRe.

See the MRTutorial(Bath), especially section 2 and the examples, for a description of some basic concepts of molecular replacement (including the search radius) and a discussion of the importance of correct choices.

The Interface documentation on features from this tutorial can be found at:

Molecular Replacement Module

File Selection

The FileViewer Utility